[ceph-users] Why keep old epochs?
gfarnum at redhat.com
Sun Nov 19 01:45:24 PST 2017
On Wed, Nov 15, 2017 at 2:07 PM Bryan Henderson <bryanh at giraffe-data.com>
> Some questions about maps and epochs:
> I see that I can control the minimum number of osdmap epochs to keep with
> "mon min osdmap epoch". Why do I care? Why would I want any but the
> osdmap, and why would the system keep more than my minimum?
> Similarly, "mon max pgmap epoch" controls the _maximum_ number of pgmap
> to keep around. I believe I need more than the most recent pgmap because I
> need to keep previous ones until all PGs that were placed according to that
> pgmap have migrated to where the current pgmap says they should be. But
> do I
> need more epochs than that, and what happens if the maximum I set is too
> to cover those necessesary old pgmaps?
Ah, you're a bit backwards here. Despite, the name, the "pgmap" is a
reporting structure about how much space PGs use, but it *does not* control
where the PGs are placed. That's entirely determined by the OSDMap.
So the cluster needs older OSDMaps to do peering and make sure it gets
up-to-date PG data when doing replication/peering/recovery/etc. The PGMaps
are only useful for reporting statistics.
I don't think you as a user have any good reason to mess around with these
at this point.
> Bryan Henderson San Jose, California
> ceph-users mailing list
> ceph-users at lists.ceph.com
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the ceph-users